let's release this thing ! #12835
let's release this thing ! #12835
Conversation
|
This looks okay on formatting fixes but there are still lingering doc content problems (#12775 (comment)) not addressed yet. |
|
Can I merge #12766 before this one, or should we wait? |
|
#12766 is a bug fix and can happen any time. |
jakebolewski
force-pushed
the
jcb/docfixes
branch
2 times, most recently
from
7264b1b to
a85589c
Compare
jakebolewski
force-pushed
the
jcb/docfixes
branch
from
d2818ba to
866241a
Compare
|
So, when Tony said "lingering docs problem", I'm not sure we realized how bad it was. In particular :
A lot of this was done by hand. I think this should make us question our release policy. We basically got nothing out of this doc transition and had to rush it. I want my last 12 hours back. And I'm pretty sure Julia Computing owes us a bottle. |
|
Drinking bourbon for 12 hours straight is also not healthy. |
| sort!(fd.order, lt = (a,b) -> ccall(:jl_type_morespecific, Int32, (Any,Any), a, b) > 0) | ||
| if haskey(ENV, "GO_AWAY") && haskey(fd.meta, sig) # temporary | ||
| error("Adding doc to $f($sig) forbidden GO AWAY") | ||
| end |
There was a problem hiding this comment.
It does look like you were having fun.
jakebolewski
force-pushed
the
jcb/docfixes
branch
from
17c17a8 to
3dcb9e3
Compare
| .. code-block:: julia | ||
|
|
||
| whos([Module,] [pattern::Regex]) | ||
| whos([io,] [Module,] [pattern::Regex]) |
There was a problem hiding this comment.
This should probably be an additional signature instead of a code block.
|
Documentation is currently broken for the following methods: (Please add to this list as you discover them)
The following methods are missing synonyms/signatures (Basically @tkelman's comment in 0636e16)
The following methods are currently missing documentation:
Other issues (not regressions):
|
|
|
|
|
|
|
|
"Package Development" and "Memory allocation analysis" have their own link in the index page but are only sub pages. All other link in the index page have their own pages. Not sure if this inconsistency matters. |
yuyichao
force-pushed
the
jcb/docfixes
branch
from
c9d6e9d to
d420cfc
Compare
jakebolewski
force-pushed
the
jcb/docfixes
branch
from
b006adf to
5b8cb58
Compare
let's release this thing !
|
|
|
🍸 🍸 🍸 |
|
Our heroes! |
|
🍻 🍶 🍷 |
|
Cheers to all the great work done here by some pretty dedicated people! 👏👏👏 |
|
Cheers! |
|
|
|
OK, was this the last thing holding up branching a 0.4 pre-release? |
|
Check your milestones before you take to the bottle. |
|
I'll open a separate issue for the type summary in repl help. |
|
@jakebolewski #11536 (comment) the roadmap for 0.4 was closed, and no other was reopened, so, how do I check the milestones??? |
|
Thanks! I hadn't seen that feature of GitHub before, had just been following the 0.4 roadmap issue, but now I'm going to start drinking, after seeing how many things still to be done, most dating from 1-3 years ago! |
|
👏 |
|
I hope I am posting this in the right place: if you are still looking for documentation issues, including the library documentation, then I would like to bring up an issue related to: JuliaCollections/DataStructures.jl#114 and the work of @mbauman. The documentation states that an IntSet holds 'nonnegative' integers, and yet trying to insert 0 into an IntSet causes a deprecation warning in all recent masters. My own preference is to remove the deprecation warning (see the discussion mentioned above), but if the deprecation warning is not removed, then the documentation should be fixed. |
Combines the `doc(::Function)` and `doc(::DataType)` methods since they were nearly identical. Enables searching for docstrings attached to specific constructors in the same manner as is currently allowed for normal methods. Prior to this it resulted in a no method error. Fixes JuliaLang#12701 where splatted method docs were shadowing others. Although not the original cause of the issue, with the change to storing signatures as `Tuple{...}` (in JuliaLang#12835) this meant that `f(x...)` was stored as `Tuple{Vararg{Any}}` which is equal to `Tuple`. Since the default value of `sig` was `Tuple` not all matching docs were returned in the example provided in JuliaLang#12701.
Combines the `doc(::Function)` and `doc(::DataType)` methods since they were nearly identical. Enables searching for docstrings attached to specific constructors in the same manner as is currently allowed for normal methods. Prior to this it resulted in a no method error. Fixes #12701 where splatted method docs were shadowing others. Although not the original cause of the issue, with the change to storing signatures as `Tuple{...}` (in #12835) this meant that `f(x...)` was stored as `Tuple{Vararg{Any}}` which is equal to `Tuple`. Since the default value of `sig` was `Tuple` not all matching docs were returned in the example provided in #12701. (cherry picked from commit 1b1e6c6) ref #13330
Combines the `doc(::Function)` and `doc(::DataType)` methods since they were nearly identical. Enables searching for docstrings attached to specific constructors in the same manner as is currently allowed for normal methods. Prior to this it resulted in a no method error. Fixes JuliaLang#12701 where splatted method docs were shadowing others. Although not the original cause of the issue, with the change to storing signatures as `Tuple{...}` (in JuliaLang#12835) this meant that `f(x...)` was stored as `Tuple{Vararg{Any}}` which is equal to `Tuple`. Since the default value of `sig` was `Tuple` not all matching docs were returned in the example provided in JuliaLang#12701.
everyone cheer, Jake is near the finish line
Mmap.*Base.process_messages@assert@nloops)@repldocumentation for macrocalls